Join Bridge Winners
All comments by Craig Zastera
You are ignoring the author of this comment. Click to temporarily show the comment.
This hand will likely require room to investigate (and even that might not be enough), so I would start with 2NT to establish a GF in spades as quickly and cheaply as possible and hope to gain some useful info via partner's rebid (BTW, in our version of Swedish Jacoby, partner's jump rebid of 4 would show a 5 card (club) suit headed by 2/3 top honors in a hand with not too much overall extra values).
April 10, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Your 4 next round almost certainly will not be interpreted as shortness in clubs. Most likely (depending on what partner rebids over 2NT), your 4 follow-up will simply be interpreted as a cue-bid, perhaps one with serious slam interest if you also had a (non-serious) 3NT call available.

However, after starting with 2NT, it is likely that I will still be able to bid *5* next time as an Exclusion Ask should I decide that is the best option (BTW, it is not clear to me whether an immediate jump to 5 over partner's 1 opening would even be Exclusion unless we have a clear agreement to that effect).
April 10, 2018
Craig Zastera edited this comment April 10, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I play immediate 3 level jump shifts (e.g. 1-3) as natural and invitational and would probably have chosen that call with this hand (although it's not ideal because of the good four card second suit).
But without that agreement, why wouldn't this be a delayed jump to 3 showing a natural invite? 2 seems out of the question as that is almost a demand for partner to pass (i.e. a weak hand with long hearts).
The alternative of 2NT (OK on strength) seems wrong as we will routinely miss 6-2 heart fits (and even 6-3 heart fits if partner is too weak to commit to game).
April 10, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
It seems to me that if a pass over their (4) by partner would not have been forcing (a questionable agreement), then his double ought to just show “convertible” extra values as he needs some way to invite you to bid on without committing to playing at the 5 level.
Thus, I do not believe that if pass of (4) is non-forcing that the double of 4 can/should be defined as “pure penalties”, i.e. showing a trump stack with no interest in our side bidding higher.

However, if passing (4) by partner would have been forcing, then *that* would be the way he would invite your co-operation in deciding whether to double them or bid on. In that case, his double would be a strong statement that he wants to defend (although you can over-ride even such a double if you think your hand is sufficiently unusually offensive in nature).
April 10, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
If I had opened this hand 3 (which I wouldn't at this vulnerability), I absolutely would *NOT* remove partner's 3NT to 4. That is just a WAG (wild assed guess).
There are plenty of hands partner might hold where 3NT
would be making opposite this hand. If I think this hand is acceptable for 3 in our partnership, I have no reason to pull partner's 3NT (particularly playing IMPs).
Now, if 3NT gets doubled, I might be justified in re-considering.
April 9, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I disagree with what Paul said. I see no reason why a partnership cannot have an agreement to play “X” and both forget simultaneously so that bidder has “Y” and partner thinks he has “Y”.
This happened the first time (and only the first time!) after we had decided to play that in auctions starting with our 2NT opening, a 3 bid by responder (direct or delayed) was always a relay to 3NT while 2N-3N was a transfer to 4.
Our auction went: 2N-3N-P.
Responder had some balanced 7 count.
Both of us would have readily acknowledged that we had an agreement that the 3NT response was an (unpassable) relay to 4. But we both happened to forget at the same time.

The only worry on the above is whether, perhaps, responder's casual tempo in “raising” 2NT to 3NT might somehow have subtly influenced me not to remember that his bid was supposed to be a relay. I'm definitely *NOT* saying that I thought “hmm, that raise to 3NT was pretty smooth–maybe he forgot”. No, what I'm saying is that while a more studied 3NT might have awakened me to remember our (new) agreement, the casual one permitted me to remain dozing.
April 9, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I voted for “fine” until I noticed the vulnerability. At unfavorable, I would not consider this bid (need 7 cold tricks at unfavorable vul).
At other vuls, 3 is fine.
April 9, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
a 5000 deal simulation of random 12 HCP balanced hands (no 5 card majors) opposite random 12 HCP balanced hands (no 5 card major and no 4=4 major suit fits) gave
3NT making on 1967 (39%) of the deals using double dummy play and defense.

This suggests that except VUL at IMPs, two 12 HCP balanced hands opposite each other (with no 8 card major fit) are probably (on average) not quite enough for 3NT.

But given that real world defense vs. real world declarer play in 3NT (especially with two 12 HCP balanced hands facing each other) will probably favor declarer (i.e. real life declarers will probably make more “unmakeable” 3NTs then they will fail in “makeable” ones) makes this a very close call (i.e. I don't think real world vs. double dummy will be quite enough to make up an 11% theoretical deficit).
April 9, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
But John,
your comment goes more to the issue of whether a partnership should agree to open “light” (i.e. lighter than traditional standards for minimum hand).
That is indeed a complex issue which I would never attempt to resolve via simulations because (as you correctly point out) there are lots of other factors in addition to accurate game and slam bidding involved–in particular the competitive benefits of getting in the first blow (it is harder to bid accurately when your side has to start with an overcall or take-out double then when it has free rein in the auction).

But the issue here is not about whether the partnership should *agree* to open light or not. It is about whether this particular hand with AK-A and 4333 shape justifies an “upgrade” to opening bid status in a partnership whose normal minimum for opening balanced hands is more traditional (e.g. a decent 12 HCPs, perhaps “Rule of 22” or such). I attempt to argue that such an upgrade is not warrented.

I will say, though, that if a partnership does agree to open “light”, that must necessarily affect responder's bidding in non-competitive (and competitive) auctions.
On the actual hand, for example, if it is deemed a “normal” opening bid, then responder certainly cannot be forcing to game with balanced 12 counts and perhaps should not even be inviting game with such!
April 9, 2018
Craig Zastera edited this comment April 9, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
The claims that East's hand is a mandatory opening are overblown.

I am not debating the merits of “light initial action” systems vs. “sound initial action” ones.

Rather, I am saying if the partnership has already decided to play some verison of “sound initial action”, e.g. “Rule of 22” to define minimum standard for opening balanced hands (and, e.g. 1NT rebids showing 12-14 HCPs), then the given
East hand may well not be worth a (1st chair) opening bid.

Here is just one data point to support my argument.

I put the given East hand (I specified the “x”s to produce
specifically A73-862-T94-AK52) opposite randomly chosen balanced hands with *12* HCPs and no 5 card major (to avoid the issue of playing 4M rather than 3NT).
To me, these represent (minimum) hands that would typically force to game (generally 3NT) after 1st chair partner shows a minimum range balanced opening hand if playing methods where (1st and 2nd) chair openings promise sound values (e.g. the popular “Rule of 22” guideline).

I did a 5000 deal simulation with the given hand playing 3NT.
How often do you suppose 3NT was makeable (double dummy)?
The answer is on only 571 (11%) of these deals (!)
Even *2NT* was makeable on only 2616 (52%) of these deals.

This suggests that the given hand may well not justify “upgrading” to opening bid status playing methods where balanced minimums are expected to be at least reasonable 12 counts, at least when considering whether partner is likely to make the right decisions when he holds minimum GF values (and also with a balanced invite).
April 8, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
My (strong) opinion on this is that any run-out system in which it is impossible to play exactly 1NX (not XX) is a bad run-out system. It may not be making, but it will not be uncommon that any attempt to run to a better contract (which may well not exist) will result in a *worse* one.
As the 1NT overcaller, if my LHO doubles and partner passes, I generally would not run unless I have a 5(+) card suit. I suppose there might be exceptions with a very concentrated 4=4.
April 8, 2018
Craig Zastera edited this comment April 8, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Henry,
I fully understand your point. But suppose intervenor has 19 or so HCPs, a decent 5 card suit, and, say, 3 card support for the other unbid suits, perhaps:
AQTxx-xx-AQx-AKx
and his RHO opens (1).
If he thinks (as I do) that this is perhaps a bit too strong for a 1 overcall, he will start with a double. When his partner advances 2m, I believe he is best advised to continue with 2 rather than the alternative 2 (which strongly suggests 3 card support and 19+ HCPs).
That is, although double-then-cuebid is *a* way to show a strong hand with 3 card support, when doubler also has a good 5 card suit (particularly a major) in addition to said 3 card support, I think he is better advised to show the major. And, I would contend, such a hand is the prototype for double then show a new suit rather than the very problematic example you give.

Obviously, this ambiguity can create problems when advancer has a weak hand with a 6 card suit as in your example.
But on your exact example, with advancer holding both a doubleton spade and a very weak hand, I think his best choice (clearly) would be to pass 2. But I know that you could give him a stiff spade and then he probably would have to gamble 3 and hope for the best.
April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Henry,
A difficult problem well worthy of “Master Solvers' Club.”
Some might overcall 1, but with 19 HCPs, many would feel this is above the upper limit.

Too strong for a direct 1NT overcall (not to mention the unbalanced shape) in my view.

If the above are rejected, then one must start with a “double.” I'd guess that would be the majority view.
When CHO advances 2, one could continue with 2 or 2NT.
(I would not consider 2 as I believe that should strongly suggest exactly 3 card support with 19+ HCPs, but perhaps some would be OK with 2).
I think it is a close decision between these two choices.
I think I would try 2.
April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Assuming you are talking about “theoretical dice”, i.e. ones which randomly and with equal probability show one of
{1 .. 6} on each roll, then a sequence of rolls is more likely to end with <1, 2> then with <6,6>.
This is easy to see without any mathamatics (to calculate the expected number of rolls to reach either ending is more difficult) by considering the following:
(a) how many ways can the sequence end after two rolls:
answer: obviously, one for each–<1, 2> and <6,6>
So, so far, both endings are equal.

(b) how many ways can the sequence end after *three* rolls:
answer:
for 6,6: *five*: <x, 6, 6> where x is any of 1 .. 5.
for 1,2: *six*: <x, 1, 2> where x is any of 1 .. 6.

Already this is enough to show that ending in 1, 2 will, on average, occur after fewer rolls than ending on 6,6, but of course one could continue this analysis for 4, 5, 6, … length sequences.
April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
In my view (and I think this is fairly mainstream), the double followed by a (non-jump) new suit implies 5 cards in the suit with playability in other strains.
With super strong (say at least 16+ HCPs, perhaps more for some who accept very strong simple overcalls), one would tend to double and then jump in a new suit (assuming not playing intermediate or strong immediate jump overcalls).

This is not 100% as there are surely hands where level, vulnerability, suit quality, strength, etc. would discourage one from doubling and jumping with a strong 1-suiter, but I think the *expectation* for double followed by a suit (non-jump) should be a 5 card suit, 18+ HCPs, and playability in other suits if advancer has 5+.
April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
If you removed, say, the minor suit queens, then a 1 overcall (not 1NT) would be clear. So here, doesn't double followed by 2 show the same with extra strength?
Further, this sequence suggests playability in clubs better than a 2NT follow-up (which might be a doubleton club).
April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think the overwhelming majority plays Jacoby 2NT exactly as stated.
However, I don't because I don't like giving away valuable information to the opponents when our side has too few values for slam. Hence, I play that opener rebids 3 with all minimums, 3 with extras and no shortess, and 3/3/3NT with extras and short //OM respectively.
Over opener's 3 (all minimums), responder with sufficient extras can bid 3 ask for shortness–then opener's 3/3/3N are as above. With no shortness, he cue-bids cheapest control (or 4M with none).
April 7, 2018
Craig Zastera edited this comment April 7, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think some of this analysis depends on what opener's (rather odd IMO) redouble means.
In my partnerships, opener has a whole host of specialized game tries available–help suit, short suit, general strength, trump–just as he would have had RHO passed instead of doubling.

I would assume that in most partnerships, opener would still have whatever arsenal of game tries the partnership deems useful after 1M-2M in an uncontested auction.

So what does the redouble mean? Assuming all sorts of game tries are already available via other calls, one might suppose that the redouble therefore shows (slightly?) less than enough for a game try. Perhaps a balanced 15? But many would have opened 1NT with such a hand, so meaning of “redouble” here is also influenced by partnership style w.r.t 1NT openings with a 5 card major.

Maybe this redouble should show a balanced 17-18 HCPs where opener was originally planning a 2NT rebid? But then, over the raise, with that good a hand he might just go ahead and make his normal game try (but if “redouble” is clearly defined as this in the partnership, it might be a better choice as it retains the chance of defending when the opponents pull).

Anyway, before one can intelligently discuss what responder's calls should mean (i.e. is double “maximal” or just penalties, would his pass be forcing, etc.) it must surely be necessary first to have a very specific definition of what opener's redouble shows as contrasted with other calls (e.g. various game tries) he might have chosen instead.
April 6, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I always assume matchpoints (since that is what I mostly play) unless the problem specifically says otherwise.

There are many, many sequences (both competitive and not) where the best definition of a call is probably different at matchpoints vs. IMPs, and all too few partnerships seem to change their methods to account for this.
April 5, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think a reasonable case (with examples) can be made for either take-out or penalty interpretation. Since our partnership generally plays low-level competitive doubles as take-out whenever that makes any sense, I would stick with that here.
As others have shown, West can easily have a hand where TO makes sense, e.g. KJx-Ax-KQxx-KQxx. With a balanced hand in the NT overcall range, it is often better to start with 1NT to show the hand type and strength rather than with a take-out double (particularly with only 3 in the other major).

And while it is easy to construct examples where West can beat 2 in his own hand (pretty much necessary for a penalty double to make sense here), how likely is he to hold such a hand when South has freely bid 2 after hearing that his LHO has a strong hand with heart stopper(s)?

Further, even with TO interpretation, since East knows West has a (near) max 1NT overcall with at least one heart stopper, he is free to convert when appropriate.
April 5, 2018
.

Bottom Home Top